After backlash, Superhuman says authors can now opt out of Grammarly's Expert Review feature by emailing a dedicated address
No apology, no overall change in direction. … They did the same to my boss Nilay Patel, my colleagues David Pierce and Tom Warren …
Superhuman's Grammarly offers an AI-powered tool called “Expert Review” that simulates feedback from famous writers, dead and living, without their permission
The tool, offered by the recently-rebranded company Superhuman, gives feedback based on the work of famous dead and living writers—without their permission.
Poetiq, which leverages existing LLMs to create “expert agents” for specific tasks, and spent just $40K to achieve high ARC-AGI-2 scores, raised a $45.8M seed
Poetiq, which leverages existing LLMs to create “expert agents” for specific tasks, and spent just $40K to achieve high ARC-AGI-2 scores, raised a $45.8M seed
Poetiq, a less-than-one-year-old A.I. startup just crushed the ARC A.G.I. benchmark, beating Anthropic and Google with only six people and $40,000.
Yann LeCun admits that Llama 4's “results were fudged a little bit”, and that the team used different models for different benchmarks to give better results
The interview took place in a great restaurant in Paris: Yannick Alléno's Pavyllon. … Bluesky: Rob Delaney / @robdelaney : 💩💩💩 [embedded post] SE Gyges / @segyges : this puts a bunch of us in the awkw...
A reflection on AI advances in the past decade and how scaling and time-horizon trends might point to far greater capabilities in the decade ahead
Zhengdong Wang : X: @zavaindar , @sparshagarwall , @justinbullock14 , @theamberyang , @firstadopter , @curl_justin , @pracheeac , @samuelalbanie , @sebkrier , @basavasagar18 , @zhengdongwang , and @s...
Pennsylvania's Supreme Court rules that police can get Google search data without a warrant; an expert warns it may encourage warrantless searches nationwide
Getting a warrant for these was already trivial. [embedded post] @toad.city : Experts warn that if violating people's privacy is already deeply ingrained in the culture of policing and you make it ea...
Xiaomi releases MiMo-V2-Flash, an open-weight MoE model with 309B total and 15B active parameters, saying it excels in reasoning, coding, and agentic scenarios
improve math, break coding. Enhance reasoning, hurt safety. ✅ Solution: Train specialized expert [image] Elie / @eliebakouch : wow, this looks like a very solid open model by Xiaomi, competing with K2...
Zoom says its “federated AI” model, combining its SLM with open- and closed-source models, got 48.1% on Humanity's Last Exam vs. 45.8% for Gemini 3 Pro w/ tools
outperforming other AI models with a 2.3% jump over the previous SOTA. ✨ HLE is one of the most rigorous tests in AI, built to measure real expert-level knowledge and deep reasoning across [image] Xue...
Researchers: OpenAI's o1 analyzes languages as well as a human expert, including inferring the phonological rules of made-up languages without prior knowledge
If language is what makes us human, what does it mean now that large language models have gained “metalinguistic” abilities?